Freitag, 11. Mai 2012

Colours in genetics

In all the joy and appreciation for genetic comparison there had always been left an annoying rest about the used color sets, changing each time. I tried to imagine a unified system, which could help to identify a certain position, tying specific components to visual avatars. No matter, what representants we choose, a genetic neighbourhood will create similar colours. We could do maps for specific reference populations and see, what kind of distribution they produce.
The first idea was to head for the triangle structure of the "Global similarity" tools of 23andme and deCODEme. I tried to create a triangle set of colours, using the CMYK system and combine them:

I made a collection of 35 drawings, focussing ancient mother figures of the paleolithic, mesolithic until early neolithic period, which could be used as representants for specific ancestry components (will be presented in more detail in the next blog). To determine the colours, I did a sort of genetic bush after roughly evaluating a whole bunch of detailed BGA-plots, mostly done by David W.

After this I tried to give each avatar the appropriate colour, placing them in a map, I once developed as a background for the east-west korridor, describing the distribution of certain haplogroups.

This is a first approach, so any comments are welcome!


  1. "there had always been left an annoying rest about the used color sets, changing each time"

    Populations already change colours in your own pictures: in 1 & 2 French Basques would be the "bluest", while in 3 & 4 for some reason Orcadians are depicted as that.

    1. Orcadian should be on the most blue corner. The french Basque belong to those in a second step, that could have a varied colour, using the K-scale ( either making it lighter or darker ) But the detracted map put some colours in disorder!